Atrial fibrillation (AF), a common arrhythmia, is associated with substantially elevated risks of stroke, cognitive decline, dementia, arterial emboli, heart failure, and cardiovascular death. Existing studies of AF rely on clinical recognitio of AF when a patient presents with symptoms, but studies in patients with implanted monitoring devices such as pacemakers or defibrillators indicate that a large proportion of AF episodes produce no symptoms at all (subclinical AF). Therefore, relying on patient symptoms to identify AF seriously underestimates AF burden, defined as the proportion of monitored time that the cardiac rhythm is AF. Convenient new external ECG patch monitors now make possible extended ambulatory monitoring at reasonable cost, and permit identification of subclinical AF and estimation of AF burden. However, there is little information about the distribution or predictors of subclinical AF or AF burden; their associations with brain structure and function or with clinical cardiovascular events; or the association of left atrial or ventricular function or structure with subclinical AF and AF burden. What little evidence is available from studies in patients with implanted devices suggests that AF complications occur whether or not patients are experiencing symptoms with their AF. We propose to conduct a study of AF and AF burden in relation to cerebral and cardiac structure and function and cardiovascular events in the Multi-Ethnic Study of Atherosclerosis (MESA). We will recruit a total of 1500 MESA participants from all 6 Field Centers using a cohort design: 1) 300 participants with clinically-recognized AF during previous MESA follow-up; 2) 450 participants at high risk for AF based on NT-proBNP level drawn in 2010-2012 and a published AF risk score, and 3) a random sample of 750 participants. The 1500 participants will have two 14-day ECG monitoring episodes in 2016-2017, and consenting participants (estimated n=1350) will have a brain MRI one to two years later. We will examine the presence of AF and AF burden from the ambulatory ECG monitoring in relation to measures from the brain MRI done one to two years after ECG monitoring, and in relation to cardiovascular events during MESA events follow-up. We will also examine measures from the 2010-2012 cardiac MRI, as well as participant characteristics and biomarkers, in relation to the presence of AF and AF burden on the monitor. Information from our study will help physicians and researchers decide on the utility of measuring subclinical AF and AF burden as new ways of quantifying AF. Knowledge about cardiac and brain abnormalities and cardiovascular events associated with the spectrum of AF burden may have implications for clinical care, including decisions about therapy to reduce the risk of stroke and heart failure and whether follow-up monitoring is warranted for patients who have had a single first episode of AF.